Bio-inspired catalytic imine reduction by rhodium complexes with tethered hantzsch pyridinium groups: Evidence for direct hydride transfer from dihydropyridine to metal-activated substrate

Article abstract of DOI:10.1002/anie.201210086

Inspired by Nature: A conceptually new design for a catalyst, combining a metal center abutted to an organic hydride donor, is demonstrated for the formate-driven transfer hydrogenation of imines under ambient conditions. A key step, transfer of hydride from the organohydride donor to the metal-polarized substrate, mirrors that in metallo-(de)hydrogenase enzymes.

Full text of DOI:10.1002/anie.201210086

Angewandte
Chemie
DOI: 10.1002/anie.201210086
Biomimetic Transfer Hydrogenation
Bio-Inspired Catalytic Imine Reduction by Rhodium Complexes with
Tethered Hantzsch Pyridinium Groups: Evidence for Direct Hydride
Transfer from Dihydropyridine to Metal-Activated Substrate**
Alex McSkimming, Mohan M. Bhadbhade, and Stephen B. Colbran*
Herein, we report a conceptually new approach to the ester, such as 2,6-dimethyl-1,4-dihydropyridine-3,5-dicarbox-
catalytic reduction of unsaturated substrates, demonstrated ylate (HEH), as the OHD center, because of their wide use as
for imine hydrogenation, based on mimicry of biological mimics for NAD(P)H in transfer hydrogenations of unsatu-
processes[1] in which hydride is directly transferred from rated substrates, such as imines or enones.[10,11] Typically,
dihydronicotinamide adenine dinucleotide (phosphate) a Brønsted or Lewis acid catalyst is required to activate the
(NAD(P)H) cofactor to an enzyme-activated substrate. substrate and to organize it and HEH for stereoselective
NAD(P)H is Natureꢀs hydride carrier.[2,3] In many (de)hy- hydride transfer.[10,11] Of note here is the recent disclosure by
drogenase enzymes that catalyze direct hydride transfer to/ Zhou et al. on ruthenium-complex-catalyzed HEH regener-
from NAD(P)+/NAD(P)H, the substrate is polarized and thus ation using dihydrogen at high pressures for the asymmetric
activated by binding to a metal ion.[4,5] Classic examples are transfer hydrogenation of cyclic oximines catalyzed by
alcohol dehydrogenases (Zn2+ active site)[4] and acetohydroxy a chiral phosphoric acid.[12] We believed that the flexible
acid isomeroreductase hydrogenases (with an (Mg2+
)
2
or electron source and the mild reduction conditions for
(Mn2+)2 active site).[5]
[Cp*RhIII(NN)L]n+ catalysis of OHD regeneration would
Our aim in this research was to prepare and test a new circumvent the demand for dihydrogen pressure and lead to
design for a homogeneous catalyst in which an unnatural a more convenient, greener, process. Moreover, substrate
organo-transition-metal center is tethered to an organohy- coordination and activation at the metal center would obviate
dride donor (OHD). The design incorporates the main the need for an expensive phosphoric acid catalyst.
features of an (de)hydrogenase enzyme and its NAD(P)H
To test our ideas, three complexes were synthesized: 1 and
cofactor into one molecule. We envisaged that the close 2, which have ortho- or meta-phenyl-bridged pyridinium
proximity of cofacial, linked metal and OHD centers would (HE+) and [Cp*Rh(NN)Cl]+ centers, respectively, and 3,[13]
facilitate both regeneration of the OHD through the inter- which lacks an HE+ substituent, as a control (Figure 1). The
mediacy of metallo-hydride species and the rapid transfer
hydride from the OHD to a metal-bound, and thus activated,
unsaturated substrate.
We targeted a [Cp*RhIII(NN)L]n+ (NN = diimine; L =
halido, n = 1; L = solvato co-ligand, n = 2) complex, as these
are the most commonly used catalysts for regeneration of
NAD(P)H from NAD(P)+.[6] Electrolytic reduction of
[Cp*RhIII(NN)L]n+ affords the corresponding RhI complex,
which is rapidly protonated at low pH to give the active
hydrido–RhIII species for hydride transfer to NAD(P)+.[6–9]
Figure 1. Metal complexes examined as catalysts in this study.
Conveniently, catalytic regeneration of OHDs using
[Cp*RhIII(NN)L]n+ can be driven directly by electricity, by
light and a photosensitizer, or by renewable chemical N-methyl derivatives were targeted because reduction of the
reductants, such as formate.[7–9] We employed a Hantzsch N-methylpyridinium cation is easier than reduction of the
neutral pyridine conjugate of a non-N-methylated dihydro-
pyridine. The new ligands and their rhodium complexes were
synthesized using well-established methods (see the Support-
ing Information for details). Of note, 1 and 2 were made by
selective oxidation of the corresponding N-methyl dihydro-
[*] A. McSkimming, Prof. S. B. Colbran
School of Chemistry, University of New South Wales
Sydney, NSW 2052 (Australia)
E-mail: s.colbran@unsw.edu.au
pyridine-substituted complexes, 1-H and 2-H, with
Dr. M. M. Bhadbhade
Mark Wainwright Analytical Centre, University of New South Wales
Sydney, NSW 2052 (Australia)
(TEMPO)[BF4].[14]
A
1D-gradient-enhanced NOESY
1H NMR experiment supports a structure for 1-H in which
the HEH-substituent and the chlorido ligand are oriented
away from each other on opposite sides of the pyridylimine
ligand (see the Supporting Information). UV/Vis spectra of
1-H and 2-H exhibit broad bands from the HEH moiety at 324
and 336 nm, respectively (this was also observed for the free
[**] This work was supported by ARC Discovery Projects
(DP1301033514). We thank the Australian Synchrotron for access to
crystallography facilities.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201210086.
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3411

.
Angewandte
Communications
ligand precursors; see the Supporting Information); these
values are typical for Hantzsch 1,4-dihydropyridines.[15] In
contrast, there are no bands above 300 nm in the UV/Vis
spectra of 1 and 2. In the FTIR spectra, 1 exhibits ester
carbonyl stretching bands at 1752 and 1727 cmꢀ1. These
values are appreciably higher than those for the carbonyl
peak in the FTIR spectra of 1-H at 1672 cmꢀ1, which is
consistent with the positively-charged HE+ ring in 1. All
complexes are air and moisture stable at ambient temperature
in the solid state and in acetone solution.
Next 1, 2, and 3 (control) were screened as catalysts for
the transfer hydrogenation of simple imines to the corre-
sponding amines using formate ion (1.1 equiv) as the electron
(hydride) source and formic acid (1.1 equiv) as a buffer. The
results are presented in Table 1. The use of formic acid or
formate alone gave poor results (entries 2g and 2h).
Table 1: Formate-driven transfer hydrogenations of imines using 1, 2,
and 3 as catalysts.
Figure 2 presents views of 1 and 2 from the X-ray crystal
structures of [1][PF6]2·acetone and [2][PF6]2. Both cations
adopt the expected “piano-stool” geometry. The bond lengths
and angles are all within reported ranges for similar
[Cp*Rh(NN)L]n+ complexes.[7f,13] The chlorido ligand in
1 sits roughly atop the HE+ ring and is 3.351 ꢁ from the
ring centroid (sum C+Cl van der Waals radii = 3.45 ꢁ) with
the closest Cl···C(HE+) distances being 3.318 and 3.311 ꢁ to
C14 A and C15 A, respectively. The Cl···p-cation ring inter-
action seems to be stabilizing.[16] It may be deduced from the
Entry R1
R2
R3
Yield [%]
Yield [%] Yield [%]
with 1[a]
with 2[a]
with 3[a]
1a
1b
2a
2b
2c
2d
2e
2 f
2g
2h
2i
3
OMe
H
H
4-MeC6H4
4-MeC6H4
97 (95)
94[b]
27
38[b]
38
–
60
8[b]
59
Me
>99 (98)
>99[c]
>99[d]
–
<2[c]
–
distinct 1H and 13C{H} NMR signals for the five methyl groups
48[e]
–
+
ꢀ
(70)[f]
(90)[g]
29[h]
in 1 that a similar structure, with proximate HE and Rh Cl
groups, is maintained in solution. In 2, the m-phenyl spacer
allows the HE and Rh Cl centers to orient away from each
+
ꢀ
33[i]
other, thus minimizing intramolecular steric interactions.
According to our initial proposition (see above), the crystal
structures suggest 1 should be a good transfer hydrogenation
catalyst, whereas 2 should be a poor one.
18[j]
Cl
H
H
4-MeC6H4
4-MeC6H4
98 (98)
9
5[k]
9
25
3[k]
5
4
5
6
NO2
Me
Cl
29[k]
Me 4-MeC6H4
Me 4-MeC6H4
H
H
H
89
79
92
94
<4
21
<3
36
7
Cl
nBu
nBu
Cy
8
9
OMe
Cl
<1
13
<1
30
93 (88)
[a] Yields were determined by 1H NMR spectroscopy (values in paren-
theses are yields of products isolated by chromatography) and are for
16 h reactions. Each reaction was performed under a nitrogen atmos-
phere with solvent (10 mL) and substrate (0.24 mmol), unless otherwise
noted. In all cases, no further conversion was observed upon extending
the reaction time to 36 h. [b] No AgOTf added. [c] Performed in air.
[d] Over Hg(l). [e] 1,2,6-Trimethyl-1,4-dihydropyridine-3,5-dicarboxylate
(5 mol%) added. [f] Imine not isolated, toluidine and tolualdehyde
(1.2 equiv) added; [g] Imine not isolated, toluidine and tolualdehyde
(2.2 equiv) added (5% 38 amine also isolated). [h] Only formic acid
(1.1 equiv) added. [i] Only formate (1.1 equiv) added. [j] 1-H used as
catalyst. [k] MeOH (15 mL) used for solubility reasons. Cy=cyclohexyl.
From the results, 1 is by far the superior catalyst; with this
catalyst the hydrogenations proceeded smoothly for most
N-alkyl/aryl aldimines and ketimines to afford the corre-
sponding amine, typically in > 90% yield. The considerably
lower conversion of the electron-deficient 4-nitrophenylimine
(Table 1, entry 4), which by thermodynamic arguments
should be the easiest to reduce, suggests that an inner-
sphere reduction mechanism, in which the imine binds to the
metal center, is operative. Initial optimization studies
revealed that increasing concentrations of formate inhibit
the reaction (see the Supporting Information), which is
consistent with formate competively binding the metal
center over the imine substrate. In contrast to 1, 2 performed
extremely poorly under identical conditions and, in most
cases, was out performed by the control catalyst. In other
Figure 2. Views of a) 1 and b) 2 with thermal ellipsoids set at 50%
(H atoms omitted for clarity). Selected distances [ꢀ] and angles [8] for
1: average Rh–C(Cp*) 2.164(4), Rh–N1 2.102(3), Rh–N2 2.175(3), Rh–
Cl 2.3937(9), Rh···C13A 4.222; N1-Rh-N2 76.16(12), N1-Rh-Cl 85.94(8),
N2-Rh-Cl 91.85(8). Selected distances [ꢀ] and angles [8] for 2: average
Rh–C(Cp*) 2.164(3), Rh–N1 2.127(3), Rh–N2 2.115(3), Rh–Cl
2.4017(10), Rh···C13A 6.518; N1-Rh-N2 76.25(10), N1-Rh-Cl 88.01(8),
N2-Rh-Cl 88.49(7).
3412
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416

Angewandte
Chemie
words, the HE+ group in 2 actually acts to inhibit imine
reduction.
(Table 1, entry 2) into 28 amine were measured. This gave
a TOF8(295 K) of (66 ꢁ 24) hꢀ1 for 1 and (3.3 ꢁ 0.5) hꢀ1 for 3
(reactions run three times). Although higher TOF8s have
been reported for imine transfer hydrogenation reactions,[19]
these results clearly demonstrate the superiority of 1 com-
pared to 3. The total turnover number (TTON) of 1 for the
reduction of imine 2 was determined to be 179 ꢁ 28 (reactions
run five times with catalyst loadings of 0.1–0.5 mol%). This
TTON value for 1 implies moderate catalyst stability under
the conditions employed, and suggests ca. 1 mol% catalyst is
required to achieve near-quantitative conversions.
A number of additional reactions offer insight into the
catalysis. First, the addition of silver triflate (to sequester the
chlorido ligand and thus ensure rapid coordination of formate
to the rhodium center) was not necessary for imine hydro-
genations using 1 (Table 1, entry 1a vs. 1b). However, the
absence of silver triflate stifled the reaction when using the
control catalyst 3, which suggests that the chlorido ligand is
less labile in this complex. This is likely due to the absence of
labilizing intramolecular steric interactions.
Performing the reaction in an open flask under air
(Table 1, entry 2b) did not hinder reactions using 1. However,
air almost completely halted reactions using the control
catalyst 3.
The electrochemistry of 1, 2, and 3 is revealing. Cyclic
voltammograms (CVs) of each complex were recorded and
examples are presented in Figure 3. The primary reduction
Metal–hydride complexes of the general formula
[Cp*Rh(NN)H]+ are extremely air sensitive.[7b,d,9] The low
catalytic activity of 3 in air most likely results from instability
of the hydrido–RhIII intermediate generated upon formate
addition. The absence of air-sensitivity in transfer hydro-
genation reactions using 1 as catalyst suggests that a hydrido–
RhIII intermediate does not form, or is very short lived.
Complex 1-H was also screened for catalytic activity.
Remarkably, this complex performed poorly (entry 2i), even
below the activity of control complex 3. Thus, 1-H can not be
an intermediate in the transfer hydrogenation of imines
catalyzed by 1 (even though 1 is the product of hydride
abstraction from 1-H, see above). Likewise, attempts to
reduce imines stoichiometrically with either 1-H+AgOTf or
2-H+AgOTf produced no amine.
We considered the possibility that 1 decomposes (because
it is so sterically hindered) into a colloidal suspension of
rhodium nanoparticles that catalyze imine hydrogenation.
This possibility was ruled out by performing the reaction over
rapidly stirred mercury (entry 2c); the mercury had no effect,
therefore the catalysis is homogenous.[17] The addition of
dimethyl-1,2,6-trimethyl-1,4-dihydropyridine-3,5-dicarboxy-
late to reactions catalyzed by 3 provided no increase in yield
(entry 2d). That is, the multi-component system containing
a metal catalyst and an HEH proved ineffective (as was 2 with
linked, but isolated, metal and HE+ centers).
Figure 3. a) Cyclic voltammograms of 1 (c), 2 (c), and 3 (a).
b) Cyclic voltammograms of 1 before (a) and after (c) addition
of 50% (v/v) TEOA/TEOA·HBF4 (0.1m, pH 7) buffer in MeOH. c) As
for (b), but with 2. Conditions: [(nBu)4N][PF6] (0.1m) in acetonitrile at
298 K; glassy carbon mini-disk (1.0 mm diameter) working electrode;
scan rate=300 mVsꢀ1 for (a) and 100 mVsꢀ1 for (b) and (c);
An attempt to bypass prior synthesis of the imine by direct
reductive amination catalyzed by 1 met with some success.
Reaction of 4-toluidine and 4-tolualdehyde (1.2 equiv) using
the standard conditions gave the 28 amine in moderate yield
(70%; entry 2e) with some toluidine remaining unreacted.
Reaction of 4-toluidine with more 4-tolualdehyde (2.2 equiv)
under the standard conditions using 1 as catalyst increased the
yield of isolated 28 amine to 90% (entry 2f), with ca. 5% of
the tertiary amine also isolated. These results suggest that the
1-catalyzed reduction of aldehydes is slow relative to imine
formation. Reduction of 4-chlorobenzaldehyde using the
standard conditions listed in Table 1 gave ca. 76% of
4-chlorobenzyl alcohol for 1 and 8% for 3. Attempted
reduction of 4-chlorobenzophenone gave no conversion
using 1 or 3. Likewise, attempted reductive aminations of
aliphatic ketones were unsuccessful.
E1/2(ferrocenium/ferrocene)= +0.47 V (and is +0.630 V vs. NHE).
process (at ꢀ0.63 V for 1, at ꢀ0.66 V for 2 and at ꢀ0.72 V for
3) is attributed to the two-electron reduction of the rhodium
ꢀ
center concerted with scission of the Rh Cl bond [Eq. (1);
L = Clꢀ], which characterizes reduction of such RhIII com-
plexes.[9] For 2 and 3, the reverse sweep shows an anodic peak
for the reverse processes [Eq. (2)]. The cathodic and anodic
processes [Eqs (1) and (2)], are not an electrochemical
couple, because a structural change accompanies the change
from RhIII to RhI (ECEC mechanism: the initial one-electron
reduction triggers diffusion-limited disproportionation of the
To glean the standard turnover frequency (TOF8)[18] of
1 versus 3, the times taken to convert ca. 50% of ditolylimine
intermediate RhII species with concomitant Rh L bond
I
ꢀ
scission).[9]
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3413

.
Angewandte
Communications
III
I
I
þ
III
½Cp Rh ðNNÞLꢂnþ þ 2 eꢀ ! ½Cp Rh ðNNÞꢂ þ L
½Cp Rh ðNNÞꢂ þ H ! ½Cp Rh ðNNÞHꢂþ
*
*
*
*
ð1Þ
ð2Þ
ð4Þ
ð5Þ
ð6Þ
I
III
III
I
ꢀ
½Cp Rh ðNNÞꢂ þ L ! ½Cp Rh ðNNÞLꢂnþ þ 2 eꢀ
½Cp Rh ðNNÞHꢂþ þ 2 eꢀ ! ½Cp Rh ðNNÞꢂ þ H
*
*
*
*
III
III
½Cp Rh ðNNÞHꢂþ þ Hþ þ L ! ½Cp Rh ðNNÞLꢂnþ þ H2
*
*
CVs of 1 and 2 also show a second reduction couple with
a
c
poor chemical reversibility (ip /ip = ca. 0.6 for both com-
plexes) at ca. ꢀ1.0 V, which corresponds to the one-electron
reduction of the HE+ group to the neutral radical (Eq. (3)],
which is unstable and consumed by radical coupling on the
CV (ms) timescale.[20] The weak Nernstian couple at ꢀ1.4 V
observed for 1 and 2, but not for 3, is attributed to the
reduction of the Hantzsch dihydropyridine dimer formed by
coupling of the radical from the previous HE+-centered
reduction.[20]
The CV response of 1 to the buffer is remarkable and
important for a number of reasons. First, the marked drop in
current for the reduction of HE+ after addition of the buffer
indicates that much of the HE+ substituent has been reduced
before the peak for its reduction is swept through. This can be
due only to rapid hydride transfer in the hydrido–RhIII species
(C, Scheme 1). That is, transfer of hydride from the metal to
the HE+ group is rapid on the CV (ms) timescale, which
explains the insensitivity of the reactions catalyzed by 1 to air
(see above). UV/Vis spectroelectrochemical analysis shows
that under protic conditions the primary (Rh-centered)
reduction cleanly produces a species with a HEH center
(lmax = 354 nm; see the Supporting Information). Second,
with pH 7 buffer, CVs of 1 show rapid formation of the
hydrido–RhIII complex [Eqs (4) and (5)], whereas those of 2
and 3 do not. Protonation of the RhI species from 1, therefore,
must be assisted by the proximate HE+ group. Exactly how is
not clear, but our hypothesis, and the basis for a theoretical
study now underway, is that the ester groups are involved.
Based on the available evidence, we propose the tentative
catalytic cycle depicted in Scheme 1 for the formate-driven
reduction of imines catalyzed by 1. Substitution of A (itself
obtained from through dissociation or Ag+ abstraction of the
chlorido ligand in 1) by formate ion affords B. Importantly,
electrostatic effects ensure formate ion attacks Rh syn to the
HE+. Hydride transfer from the formato ligand to the metal
HEþ þ eꢀ ! HE
ð3Þ
C
Notably, in the CVs of 1, the anodic peak for the re-
oxidation of the RhI species shifts considerably to + 0.12 V
(indicated by a star in Figure 3a) and exhibits a marked
decrease in current. The drop in current reveals that the RhI
state is rapidly lost. Connectivity and the relative orientation
of Rh and HE+ units define the only difference between 1 and
2. The different behavior upon reduction of 1, as compared to
+
ꢀ
2 and 3, must arise from the adjacent Rh Cl and HE centers
in the former complex. Therefore, the decreased current for
re-oxidation of the RhI center, suggests that the RhI center is
oxidized by the adjacent HE+ group. The shift in the re-
oxidation peak for 1 (vs. 2) of ca. 500 mV suggests increased
thermodynamic stability of the RhI state in 1. It is not clear
why, although we suspect it has to do with stabilizing
interactions between the rhodium center and the ester
groups of the HE+/HEH substituents.[7f]
then produces the RhIII H intermediate C and CO2, which
ꢀ
Figure 3b,c compares CVs of 1 and 2 before and after
addition of a pH 7 buffer. For 2 and 3 (not shown), the overall
reduction chemistry is unchanged by the buffer: only small
potential shifts and some broadening of the peaks is seen. In
contrast, the addition of pH 7 buffer
would be rapidly followed by hydride transfer to the adjacent
HE+ pyridinium group to give intermediate D with a HEH
dihydropyridine group with the C4 hydrogen pointing directly
at the RhIII center (as opposed to 1-H, which is inactive and
to 1 causes dramatic changes to the
redox behavior of the complex. The
CVs reveal lower current for the
primary reduction at ꢀ0.52 V
[Eq. (1)], and for the HE+/HEH
couple at ꢀ1.04 V, [Eq. (3)], and
the appearance of a new irreversible
peak at ꢀ0.76 V. The new peak
corresponds to reduction of the
RhIII H complex formed by rapid
ꢀ
protonation of the RhI species,
[Eqs (4) and (5)]. Catalytic evolution
of dihydrogen [Eqs (6), (1) and (4)],
may contribute to the current for this
peak. This behavior has been
thoroughly
characterized
for
[Cp*RhI(NN)]0 species under protic
conditions.[9]
Scheme 1. Possible mechanism for the reduction of imines by formate catalyzed by 1.
3414
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416

Angewandte
Chemie
has the C4 hydrogen orientated away from the RhIII center;
see the Supporting Information). Alternatively, D could form
directly from B by hydride transfer from the formato ligand to
the HE+ group.[21] Coordination of an imine gives E, in which
the imine is activated for hydride attack. Coordination of
a second equivalent of formate to Rh would compete with this
metal catalyst (0.0024 mmol) and AgOTf (0.0026 mmol) were com-
bined in a Schlenk flask and dissolved in MeOH (9.0 mL). The
solution was deaerated with a steady stream of dinitrogen for ca.
10 min, at which point MeOH (1.0 mL) containing HCO2H/NaHCO2
(1:1, 0.530 mmol total) was injected. The solution was purged with
dinitrogen for more 5 min and then the flask was sealed. Aliquots
were taken at various intervals: the MeOH was removed, the residue
1
ꢀ
step and lead to a dead-end Rh H species, thus accounting for
taken up in CDCl3, and the sample analyzed by H NMR spectros-
copy. For product isolation, the residue was subjected to column
chromatography (60–200 mm silica support; ether/hexanes gradient as
eluent). For catalyst 1, AgOTf is unnecessary and the reactions may
be run in an open vessel in air.
the lower yields with more equivalents of formate (see the
Supporting Information). From E, intramolecular hydride
transfer from the HEH moiety to the metal-activated imine
affords the amido species F. Protonation of the amido ligand
in F and loss of the amine product closes the catalytic cycle.
In summary, we have demonstrated a new bio-inspired
approach to chemical reduction that operates under ambient
conditions in air (even in a teacup; see the Supporting
Information). Catalyst 1 has co-joined [Cp*RhIII(NN)Cl]+
and Hantzsch HE+/HEH centers that mimic the active site
in a metallo-(de)hydrogenase enzyme and its NAD(P)+/
NAD(P)H OHD-cofactor, respectively. Crucial to the reduc-
tion of imines by 1 are: 1) the metal-catalyzed reduction of
the HE+ group to the dihydropyridine HEH, which leaves an
open the site for imine coordination, and 2) back transfer of
hydride from the HEH group to the metal-bound and
activated imine. The results demonstrate that careful design
of the multifunctional OHD-ligand is essential: the metal
center and HE+/HEH centers must abut one another for
rapid intramolecular hydride transfer. We believe that this
biomimetic design will have wide application in chemical
reduction. For example, although this example is driven by
formate ion, which is a renewable reductant,[22] it is also likely
that these reductions can be driven electrochemically or
photochemically.[7,8] We note that 1 is axially chiral. It remains
to be seen whether the enantiomers of 1 or other chiral metal–
OHD conjugates are active for asymmetric reduction. Also,
the new catalyst design is not restricted to a Rh center or to
a Hantzsch dihydropyridine OHD. There is a wide range of
metal and OHD centers that may be co-joined and we
anticipate some may be active for the reduction of more
difficult substrates (including, perhaps, carbon dioxide and
dinitrogen).
Received: December 18, 2012
Published online: February 26, 2013
Keywords: biomimetic catalysis · organo-hydride donor ·
.
redox-active ligands · rhodium · transfer hydrogenation
[1] V. K. K. Praneeth, M. R. Ringenberg, T. R. Ward, Angew. Chem.
2012, 124, 10374 – 10380; Angew. Chem. Int. Ed. 2012, 51, 10228 –
10234.
[2] J. Berg, J. Tymoczko, L. Stryer, Biochemistry, 6th ed., W. H.
Freeman and Co., New York, 2007.
[3] a) K. S. You, Crit. Rev. Biochem. 1985, 17, 313 – 451; b) N. Pollak,
C. Dçlle, M. Ziegler, Biochem. J. 2007, 402, 205 – 218.
[4] a) S. Hammes-Schiffer, S. J. Benkovic, Annu. Rev. Biochem.
2006, 75, 519 – 541; b) Z. D. Nagel, J. P. Klinman, Chem. Rev.
2006, 106, 3095 – 3118; c) D. Roston, A. Kohen, Proc. Natl. Acad.
Sci. USA 2010, 107, 9572 – 9577.
[5] a) R. Dumas, V. Biou, F. Halgand, R. Douce, R. G. Duggleby,
Acc. Chem. Res. 2001, 34, 399 – 408; b) F. Proust-De Martin, R.
Dumas, M. J. Field, J. Am. Chem. Soc. 2000, 122, 7688 – 7697.
[6] F. Hollmann, I. W. C. E. Arends, K. Buehler, ChemCatChem
2010, 2, 762 – 782.
[7] For seminal reports on NAD(P)H regeneration using
[Cp*Rh(NN)L] catalysis, see: a) E. Steckhan, S. Herrmann, R.
Ruppert, J. Thommes, C. Wandrey, Angew. Chem. 1990, 102,
445 – 447; Angew. Chem. Int. Ed. Engl. 1990, 29, 388 – 390; b) E.
Steckhan, S. Herrmann, R. Ruppert, E. Dietz, M. Frede, E.
Spika, Organometallics 1991, 10, 1568 – 1577; c) D. Westerhau-
sen, S. Herrmann, W. Hummel, E. Steckhan, Angew. Chem.
1992, 104, 1496 – 1498; Angew. Chem. Int. Ed. Engl. 1992, 31,
1529 – 1531; d) F. Hollmann, A. Schmid, E. Steckhan, Angew.
Chem. 2001, 113, 190 – 193; Angew. Chem. Int. Ed. 2001, 40, 169 –
171; e) H. C. Lo, O. Buriez, J. B. Kerr, R. H. Fish, Angew. Chem.
1999, 111, 1524 – 1527; Angew. Chem. Int. Ed. 1999, 38, 1429 –
1432; f) H. C. Lo, C. Leiva, O. Buriez, J. B. Kerr, M. M.
Olmstead, R. H. Fish, Inorg. Chem. 2001, 40, 6705 – 6716.
[8] For illustrative recent examples, see: a) H. K. Song, S. H. Lee, K.
Won, J. H. Park, J. K. Kim, H. Lee, S. J. Moon, D. K. Kim, C. B.
Park, Angew. Chem. 2008, 120, 1773 – 1776; Angew. Chem. Int.
Ed. 2008, 47, 1749 – 1752; b) F. Hildebrand, C. Kohlmann, A.
Franz, S. Lutz, Adv. Synth. Catal. 2008, 350, 909 – 918; c) S. H.
Lee, J. Ryu, D. H. Nam, C. B. Park, Chem. Commun. 2011, 47,
4643 – 4645; d) S. H. Lee, H. J. Lee, K. Won, C. B. Park, Chem.
Eur. J. 2012, 18, 5490 – 5495; e) S. H. Lee, D. H. Nam, C. B. Park,
Adv. Synth. Catal. 2009, 351, 2589 – 2594; f) C. Leiva, H. C. Lo,
R. H. Fish, J. Organomet. Chem. 2010, 695, 145 – 150.
Finally, we note that this year, 2013, is the centenary of the
award of the Nobel Prize for chemistry to Alfred Werner for
his development of the coordination theory of transition
metal complexes.[23] Wernerꢀs Nobel award was the first (and
the only one for 72 years) in inorganic chemistry. Arthur
Hantzsch was Wernerꢀs doctoral advisor, co-author of his
doctoral thesis, and friend. He was also the discoverer of the
synthetic route to the dihydropyridines that bear his name.[24]
It has only taken 100 years to combine the work of doctoral
advisor and student into one molecule. As we have demon-
strated, provided the stereochemistry is correct (something
that both Werner and Hantzsch would have appreciated and
enjoyed)[25] novel catalytic behavior, well above and beyond
that of the individual components, results.
[9] For (spectro-)electrochemical studies of [Cp*Rh(NN)L] com-
plexes, see: a) M. Ladwig, W. Kaim, J. Organomet. Chem. 1991,
419, 233 – 243; b) U. Kçlle, B. S. Kang, P. Infelta, P. Comte, M.
Gratzel, Chem. Ber. 1989, 122, 1869 – 1880; c) C. Caix, S.
Chardon-Noblat, A. Deronzier, J. C. Moutet, S. Tingry, J.
Organomet. Chem. 1997, 540, 105 – 111; d) S. Chardon-Noblat,
S. Cosnier, A. Deronzier, N. Vlachopoulos, J. Electroanal. Chem.
Experimental Section
Standard conditions for the transfer hydrogenation of imines using
HCO2H/NaHCO2 and 1, 2, and 3 as catalysts: Imine (0.240 mmol),
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3415

.
Angewandte
Communications
1993, 352, 213 – 228; e) M. Ladwig, W. Kaim, J. Organomet.
Chem. 1992, 439, 79 – 90.
[10] a) M. Rueping, J. Dufour, F. R. Schoepke, Green Chem. 2011, 13,
1084 – 1105; b) J. G. de Vries, N. Mrsic, Catal. Sci. Technol. 2011,
1, 727 – 735; c) C. Zheng, S. L. You, Chem. Soc. Rev. 2012, 41,
2498 – 2518.
[16] a) D. QuiÇonero, C. Garau, C. Rotger, A. Frontera, P. Ballester,
P. A. Costa, P. M. Deya, Angew. Chem. 2002, 114, 3539 – 3542;
Angew. Chem. Int. Ed. 2002, 41, 3389 – 3392; b) O. B. Berryman,
V. S. Bryantsev, D. P. Stay, D. W. Johnson, B. P. Hay, J. Am.
Chem. Soc. 2007, 129, 48 – 58; c) S. E. Wheeler, K. N. Houk, J.
Phys. Chem. A 2010, 114, 8658 – 8664.
[11] a) L. Simon, J. M. Goodman, J. Am. Chem. Soc. 2008, 130, 8741 –
8747; b) O. Gutierrez, R. G. Iafe, K. N. Houk, Org. Lett. 2009, 11,
4298 – 4301; c) J. W. Yang, B. List, Org. Lett. 2006, 8, 5653 – 5655;
d) S. L. Zhou, S. Fleischer, K. Junge, M. Beller, Angew. Chem.
2011, 123, 5226 – 5230; Angew. Chem. Int. Ed. 2011, 50, 5120 –
5124.
[12] a) Q. A. Chen, M. W. Chen, C. B. Yu, L. Shi, D. S. Wang, Y.
Yang, Y. G. Zhou, J. Am. Chem. Soc. 2011, 133, 16432 – 16435;
b) Q. A. Chen, K. Gao, Y. Duan, Z. S. Ye, L. Shi, Y. Yang, Y. G.
Zhou, J. Am. Chem. Soc. 2012, 134, 2442 – 2448; c) F. Shi, L.-Z.
Gong, Angew. Chem. 2012, 124, 11587 – 11589; Angew. Chem.
Int. Ed. 2012, 51, 11423 – 11425.
[17] J. A. Widegren, R. G. Finke, J. Mol. Catal. A 2003, 198, 317 – 341.
[18] S. Kozuch, J. M. L. Martin, ACS Catal. 2012, 2, 2787 – 2794.
[19] “Imino Reductions by Transfer Hydrogenation”: M. Wills in
Modern Reduction Methods (Eds.: P. Andersson, I. M. Mun-
slow), Wiley-VCH, Weinheim, 2008, chapter 11, pp. 271 – 296.
[20] F. T. Mcnamara, J. W. Nieft, J. F. Ambrose, E. S. Huyser, J. Org.
Chem. 1977, 42, 988 – 993.
[21] H. Konno, K. Sakamoto, O. Ishitani, Angew. Chem. 2000, 112,
4227 – 4229; Angew. Chem. Int. Ed. 2000, 39, 4061 – 4063.
[22] A. Boddien, D. Mellmann, F. Gꢂrtner, R. Jackstell, H. Junge, P. J.
Dyson, G. Laurenczy, R. Ludwig, M. Beller, Science 2011, 333,
1733 – 1736.
[13] P. Govindaswamy, Y. A. Mozharivskyj, M. R. Kollipara, Poly-
hedron 2005, 24, 1710 – 1716.
[14] M. Shibuya, M. Tomizawa, Y. Iwabuchi, J. Org. Chem. 2008, 73,
4750 – 4752.
[23] K. H. Ernst, F. R. W. P. Wild, O. Blacque, H. Berke, Angew.
Chem. 2011, 123, 10970 – 10977; Angew. Chem. Int. Ed. 2011, 50,
10780 – 10787.
[24] A. Hantzsch, Chem. Ber. 1881, 14, 1637 – 1638.
[15] a) R. A. Dommisse, F. C. Alderweireldt, Bull. Soc. Chim. Belg.
1973, 82, 441 – 445; b) R. F. Affeldt, R. S. Iglesias, F. S. Rodem-
busch, D. Russowsky, J. Phys. Org. Chem. 2012, 25, 769 – 777.
[25] A. Hantzsch, A. Werner, The Elements of Stereochemistry, The
Chemical Publishing Company, Easton, PA, 1901.
3416
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3411 –3416